AI and the need for purpose-built cloud infrastructure

Modern AI solutions augment human understanding, preferences, intent, and even spoken language. AI improves our knowledge and understanding by delivering faster, more informed insights that fuel transformation beyond anything previously imagined. The challenge of this rapid growth and transformation is that AI’s demand for compute power is outpacing Moore’s Law in computing advancements.

AI requires infrastructure that can meet the continually increasing compute power demands and specialized needs of AI applications and workloads, like natural language processing, robot-powered process automation, and machine learning and deep learning.

High-performance computing provides scalable solutions for AI.

To perform at today’s much higher demand levels, AI infrastructure must scale up to take advantage of single servers with multiple accelerators and scale out to combine many such servers distributed across a high-performance network.

Scale-up AI computing infrastructure combines memory from individual graphics processing units (GPUs) into a large, shared pool to tackle larger and more complex models. When united with the incredible vector-processing capabilities of the GPUs, high-speed memory pools have proven to be extremely effective at processing large multidimensional arrays of data.

With the added capability of a high-bandwidth, low-latency interconnect fabric, scale-out AI-first infrastructure can significantly accelerate time to output. This is achieved via advanced parallel communication methods, interleaving computation and communication across a vast number of compute nodes.

Cloud infrastructure purpose-built for AI

Microsoft Azure is currently the only global public cloud service provider that provides purpose-built AI supercomputers with massively scalable scale-up-and-scale-out IT infrastructure comprised of NVIDIA Quantum InfiniBand interconnected NVIDIA Ampere A100 Tensor Core GPUs. Azure Machine Learning provides enterprise-grade service for the end-to-end machine learning lifecycle, accelerating the integration of AI into workloads to drive smarter simulations and accelerate intelligent decision-making.

Scale-up-and-scale-out infrastructures powered by NVIDIA GPUs and NVIDIA Quantum InfiniBand networking rank amongst the most powerful supercomputers on the planet. Microsoft Azure placed in the top 15 of the Top500 supercomputers worldwide and currently five systems in the top 50 use Azure infrastructure with NVIDIA A100 Tensor Core GPUs. Twelve of the top twenty ranked supercomputers in the Green500 list use NVIDIA A100 Tensor Core GPUs.

**Source:** Top 500 The List: Top500 November 2022, Green500 November 2022.

This supercomputer-class AI infrastructure is accessible to researchers and developers in organizations of any size around the world and is used by customers across industry segments to meet AI’s growing computing demands. All types of AI technology, research, and applications are fulfilled, augmented, and/or accelerated with Azure’s AI-first infrastructure.

Retail and AI

A prime industry example is retail where AI-first cloud infrastructure and toolchain from Microsoft Azure featuring NVIDIA GPUs are having a significant impact. See how Everseen created a seamless shopping experience that benefits their bottom line. With a GPU-accelerated computing platform, customers can churn through models quickly and determine the best-performing model. And autonomous checkout enables retailers to provide customers with frictionless and faster shopping experiences while increasing revenue and margins. Benefits of AI-first cloud infrastructure for retail include:

Performance improvements for classical data analytics and machine learning processes at scale.
Accelerated training of machine learning algorithms. With RAPIDS with NVIDIA GPUs, retailers can use larger data sets and process them faster with more accuracy, allowing real-time reaction to shopping trends and inventory cost savings at scale.
Forecasting accuracy, resulting in cost savings from reduced out-of-stock and poorly placed inventory.

Better and faster customer checkout experience and reduced queue wait time.
Reduced shrinkage—the loss of inventory due to theft such as shoplifting or ticket switching at self-checkout lanes, which costs retailers $62 billion annually, according to the National Retail Federation.

In retail, data-driven solutions require sophisticated deep learning models—models that are much more sophisticated than those offered by machine learning alone. Deep learning also requires significantly more computing power, making optimization via an AI-first infrastructure and AI toolchain a necessity.

Learn more about purpose-built infrastructure for AI.

AI is everywhere and its application is growing rapidly. Optimized AI-first infrastructure is critical in the development and deployment of AI applications. Microsoft Azure scale-up-and scale-out infrastructure combines the power of NVIDIA GPUs and NVIDIA networking in the cloud to offer the right-sized GPU acceleration for AI applications of any scale and for organizations of any size.

With a total solution approach that combines the latest GPU architectures and software designed for the most compute-intensive AI training and inference workloads, Microsoft and NVIDIA are paving the way to go beyond exascale AI supercomputing. Learn how Azure and NVIDIA can help power your AI.